Predicting mediaplan traffic

• US 8,954,567 B1
• Filed: 06/29/2011
• Issued: 02/10/2015
• Est. Priority Date: 06/29/2011
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: Hubinette, Carl F., Skvortsov, Evgeny
• Primary Examiner(s): Hwang, Joon H
• Assistant Examiner(s): JOSHI, SURAJ M
• Application Number:
  US13/171,901
• Time in Patent Office: 1,322 Days
• Field of Search: 709/224
• US Class Current: 709/224
• CPC Class Codes: G06Q 30/0202 Market predictions or forecasting for commercial activities
• CPC Class Codes: H04N 21/24 Monitoring of processes or resources, e.g. monitoring of server load, available bandwidth, upstream requests monitoring of server performance or load G06F11/34; arrangements for observation, testing or troubleshooting for broadcast or for distribution combined with broadcast H04H20/12
• CPC Class Codes: H04N 21/25 Management operations performed by the server for facilitating the content distribution or administrating data related to end-users or client devices, e.g. end-user or client device authentication, learning user preferences for recommending movies maintenance or administration in data networks H04L41/00

Abstract

Methods, systems, and apparatus for predicting traffic data, including a method comprising: receiving data indicative of traffic data for a mediaplan during first and second time spans, the data representing numbers of unique entities that have interacted with a location during the first and second time spans; and using a function to predict third traffic data during the third time span. The function, when the arbitrary third time span does not exceed the first and second time spans, uses an exponential that includes a difference of dot products using natural logs of combinations of the time spans and first and second traffic data, divided by a difference of natural logs of the first and second time spans. Otherwise, the function uses a difference of dot products using combinations of the time spans and the first and second traffic data, divided by a difference of the first and second time spans.

Claims

1. A method comprising:

receiving, by a traffic data store, first data indicative of first traffic data for a mediaplan during a first time span and second data indicative of second traffic data for the mediaplan during a second time span, the first and second traffic data representing numbers of unique entities that have interacted with a location during, respectively, the first and second time spans;

using, by a prediction engine, a function derived from at least one of data received from a user device, the received first data, and the received second data and one of an arbitrary number of unique entities or an arbitrary third time span to predict third traffic data representing a number of unique entities interacting with the location during the third time span; and

outputting the prediction of third traffic data to the mediaplan, the mediaplan associated with at least one of an advertising campaign and a media strategy;

where the function, when the arbitrary third time span does not exceed the first and second time spans, uses an exponential that includes a difference of dot products using natural logs of combinations of the first, second and arbitrary third time spans and the first and second traffic data, divided by a difference of natural logs of the first and second time spans; and

where the function otherwise uses a difference of dot products using combinations of the first, second and arbitrary third time spans and the first and second traffic data, divided by a difference of the first and second time spans.

2. The method of claim 1 where the first and second traffic data represent measured or estimated points on a graph of time versus unique entities that have interacted with a location, where the graph includes axes in a logarithmic scale.

3. The method of claim 2 where the graph is presented in a user interface in which a user input of a value corresponding to one of the axes is used to predict a value corresponding to the other one of the axes.

4. The method of claim 1 where the first traffic data is estimated traffic data.

5. The method of claim 1 where the first traffic data is measured traffic data.

6. The method of claim 1 further comprising using the function to estimate a number of unique entities that have interacted with the location for a past arbitrary time period.

7. The method of claim 1 further comprising using the function to predict a number of unique entities that will interact with the location in a future arbitrary time period.

8. The method of claim 1 further comprising using the function to estimate a time period in which an arbitrary number of unique entities interacted with the location.

9. The method of claim 1 where unique entities are selected from the group comprising unique users, unique households, unique cookies, unique browsers, unique email addresses, and unique IP addresses.

10. The method of claim 1 where unique entities that have interacted with the location comprises unique entities that have had interactions with the location where the interactions are selected from the group comprising impressions, clicks, conversions, subscriptions, registrations, web site visits, search queries, database queries and downloads.

11. The method of claim 1 where the first and second time spans are substantially different in length.

12. The method of claim 1 where the unique entities that have interacted are determined by cookies.

13. The method of claim 1 where the unique entities that have interacted are determined by Internet Protocol (IP) addresses.

14. The method of claim 1 where the unique entities are users that have performed an action at or in association with the location.

15. The method of claim 1 where the location is a website.

16. The method of claim 1 where predicting the third data point includes predicting an amount of time required to accumulate a predetermined number of unique entities.

17. The method of claim 1 where predicting the third data point includes predicting a number of unique entities given a predetermined amount of time.

18. A method comprising:

determining first traffic data for a mediaplan during a first time span and second traffic data for the mediaplan during a second different time span, where the first and second different time spans are of substantially different lengths;

storing at least one of the first traffic data and the second traffic data in a traffic data store;

identifying an arbitrary time span over which predicted traffic data is desired to be determined;

receiving, by a prediction engine, at least one of data received from a user device, the stored first traffic data, and the second traffic data;

performing a curve fitting on the first traffic data and the second traffic data;

determining, by a prediction engine, the predicted traffic over the arbitrary time span based on the curve fitting; and

outputting the predicted traffic to the mediaplan, the mediaplan associated with at least one of an advertising campaign and a media strategy.

19. The method of claim 18 where:

performing the curve fitting on the first traffic data and the second traffic data comprises plotting, on a graph of time versus unique entity numbers, the first traffic data and the second traffic data, producing two points on the graph, where x and y axes of the graph are in a logarithmic scale; and

determining the predicted traffic over the arbitrary time span based on the curve fitting comprises;

determining a line through the two points; and

locating a point of intersection that corresponds to the arbitrary time span on the line, where a value of unique entity numbers at the point of intersection corresponds to the predicted traffic over the arbitrary time span.

20. The method of claim 18 where the mediaplan is associated with an advertising campaign.

21. The method of claim 18 where the mediaplan includes one or more websites.

22. The method of claim 18 where identifying first and second traffic data includes identifying the second traffic data and determining the first traffic data from the identified second traffic data.

23. A method comprising:

receiving traffic data at a traffic data store,

identifying, using the stored traffic data, first traffic data U_t1 for a mediaplan S during a first time span t₁ and second traffic data U_t2 for the mediaplan S during a second different time span t₂, where the first time span t₁ and the second different time span t₂ are of substantially different lengths;

identifying an arbitrary time span t over which predicted traffic data is desired to be determined;

identifying a number of unique entities U_t that visited a particular media-plan in a duration of t days;

predicting a number of unique entities U_t(S) that satisfy the mediaplan S over the arbitrary time span t including using the first traffic data U_t1 and the second traffic data U_t2 in accordance with the following formulas;

for the arbitrary time span t less than or equal to t₂ days;

24. The method of claim 23 where the mediaplan is associated with an advertising campaign.

25. The method of claim 23 where the mediaplan includes one or more websites.

26. The method of claim 23 where the unique entities are selected form the group comprising unique users, unique households, unique cookies, unique browsers, unique email addresses, and unique Internet Protocol (IP) addresses.

27. The method of claim 23 where identifying first and second traffic data includes retrieving first and second traffic data.

28. The method of claim 23 where identifying first and second traffic data includes identifying the second traffic data and determining the first traffic data from the identified second traffic data.

29. A system comprising:

one or more processors communicatively coupled to a memory;

one or more non-transitory computer-readable storage media configured to store instructions executable by said one or more processors;

a prediction engine that uses prediction models stored in the storage media and transmitted to the memory for execution by said one or more processors and user traffic data to make predictions comprising predicting a number of unique entities U_t(S) that satisfy a mediaplan S over an arbitrary time span t including using first traffic data U_t1 during a first time span t₁ and second traffic data U_t2 during a second different time span t₂, where the first time span t₁ and the second different time span t₂ are of substantially different lengths, in accordance with the following formulas;

for the arbitrary time span t less than or equal to t₂ days;

30. A computer program product tangibly embodied in a non-transitory computer-readable storage device and comprising instructions that, when executed by one or more processors, perform a method for predicting traffic for a mediaplan, the method comprising:

identifying first traffic data U_t1 for a mediaplan S during a first time span t₁ and second traffic data U_t2 for the mediaplan S during a second different time span t₂, where the first time span t₁ and the second different time span t₂ are of substantially different lengths;

identifying an arbitrary time span t over which predicted traffic data is desired to be determined;

identifying a number of unique entities U_t that visited a particular media-plan in a duration of t days; and

predicting, by a prediction engine, a number of unique entities U_t(S) that satisfy the mediaplan S over the arbitrary time span t including using the first traffic data U_t1 and the second traffic data U_t2 in accordance with the following formulas;

for the arbitrary time span t less than or equal to t₂ days;